1. Field of the Invention
This invention relates to a method of modeling reference data in an information system (IS) and an algorithm based on this modeling in order to localize deposits of reference data for an information system.
Therefore, the invention relates to the domain of software techniques.
2. Description of the Related Art
In an organization, an IS is composed of the set of all elements that contribute to the processing and circulation of information in the organization (database, application software, procedures, documentation, etc.) including the information system itself (central processing unit, peripherals, operating system, etc.). Therefore, the IS is composed of the human and technical means necessary for the storage and processing of information in an organization.
An IS is modeled in the IS design phase and it consists of creating a virtual representation of processes used by businesses in an organization and information stocks and flows providing input to these processes.
One known method called MERISE is the most widespread method used for designing an information system. It is a computer analysis method dedicated to setting up an IS. The MERISE method is based on separation of data and processing to be done in several conceptual and physical models. Among these models, the Conceptual Data Model (CDM) is used to set up a clear representation of IS data and defines functional dependences of these data between themselves. The purpose of the CDM is to formally write data that will be used by the IS. Therefore, it is an easily understandable representation of data used to write the IS using entities. An entity is the representation of a material or immaterial element that plays a role in the system that is to be described. Entities of the same type, in other words with the same definition, are grouped within a set called an entities class. Semantic links may exist between several entities. Thus, relations are defined to represent these semantic links and a relation class contains all relations of the same type (therefore that connect entities belonging to the same entity classes).
However, data modeling proposed in a Conceptual Data Model originating from the MERISE method cannot be used to translate the concept of the reference value of a data item or a reference relation between two items of data.
The concept of IS reference data relates to data produced by an application that supplies official values for these data. This concept of reference data is extremely important in an IS, particularly when it is required to create or update the urbanism of an IS. Data are often manipulated through several applications and their values can thus be modified by the different players involved in the IS. Moreover, if an IS is to remain efficient or if its services are to be improved, it must necessarily progress at the same rate as the organization develops and its strategy changes. Urbanism work then consists of defining the main lines of these developments and continuous changes, and setting up a simpler interconnection between the different systems and applications of the IS.
In this context, the concept of a reference value for a data item or a relation between two data items is very important to optimize urbanism of an IS. Each application of an IS that has reference data in its data deposit must be capable of offering these reference data to client applications. Obviously, this concept of a data offer by a supplier to customers has an influence on the functional urbanism of the IS.
The Conceptual Data Model derived from the MERISE method is incapable of representing homogenous data groups or dependence relations between these groups, although these are useful for representing the concept of reference. This method does not enable construction of a model using principles based on IS reference data and is therefore limited in terms of optimizing the functional urbanism phase of an IS.
Moreover, modeling of data proposed in a Conceptual Data Model derived from the MERISE Method is also limited in that it cannot locate deposits of reference data for new data in an IS. However, this is a very important characteristic necessary to optimize urbanism of an IS.
Thus, considering the above, one purpose of this invention is to propose a method for modeling data of an IS that overcomes the disadvantages of prior art as mentioned above, and particularly that is capable of defining a diagrammatic representation of the concept of a reference value of a dependence relation between two items of data and data themselves.
Another purpose of this invention is to propose a method for localizing deposits of reference data for new data in an IS, to support principles of the proposed modeling method, that determines which data deposit contains a reference data item, in other words the particular application responsible for the data.